Reliance Industries Chairman Mukesh Ambani joined Nvidia founder and CEO Jensen Huang for a fireside session as part of the Nvidia AI Summit India at Mumbai’s Jio World Convention Centre today. The fireside session saw the discussion around AI’s pivotal role in reshaping industries and how India is emerging as a global AI leader.
During the AGM earlier in August, Reliance Industries Ltd Chairman Mukesh Ambani’s address to shareholders reflected the conglomerate’s growing emphasis on advanced technology and innovation. Ambani’s made more than 50 references to artificial intelligence (AI) as he laid out an ambitious blueprint of deeptech moves to transform the oil-to-telecom conglomerate.
“India is very dear to the world’s computer industry and central to the IT industry,” said Jensen Huang, as he began his speech at the Summit with “Hello, Mumbai!”
“Industries are going through tectonic and seismic changes,” says Huang, adding, “There are two fundamental shifts happening at the same time that haven’t happened since 1964,” said Huang.
According to him, these changes are: The separation of hardware and application software through a layer called the OS and Moore’s law making it possible for us to continue to drive down cost.
“Today’s computer industry, the same industry where general-purpose computing has existed for 60 years, is continuing to process more and more data. Moore’s Law helped us invest in systems like the System/360, and Moore’s Law with Windows PCs undoubtedly drove the most important industry in the world. However, the scaling of CPUs has reached its limit. We cannot continue to rely solely on Moore’s Law; we must do something different or depreciation will end. Otherwise, we will experience computing inflation. We can no longer afford to do nothing in software and expect our computing experience to improve while costs decrease.”
We have democratized computer graphics, said Huang.
“Over the last 30 years, we have been on a journey to accelerate one domain of application over another. There is no such magical processor that can accelerate everything in the world. You need to reinvent the computer stack from the algorithms to the underlying architecture and connect it to applications.”
“We can no longer afford to do nothing in software and expect that our computing experience will continue to improve,” says Huang.
Every time I have come to India, I have talked about deep learning and machine learning. It’s very clear now that the world has completely changed. The first thing that happened was how we do software. Our industry is underpinned by the method with which software was done. In software 1.0, someone would write Python code, develop algorithms, apply input, and an algorithm is produced. This created one of the largest industries in the world right here in India. Coding and programming became a whole industry. However, that approach of developing software has become disruptive. It is not now coding but machine learning; using a computer to study patterns of correlation between massive amounts of data to essentially learn from it the function that predicts it. We are designing a universal function approximator.”
“The computer is now writing the software. We have gone from coding to machine learning, from developing software to creating AI, from software that prefers to run on CPUs to running neural networks that run on GPUs. We have now seen a complete reinvention of the computing stack.”
“Today, we are translating large bodies of English text to other languages, from pixels to image recognition, from images or videos to word captioning, from words to proteins used for drug discovery, from words to chemicals discovering new compounds, from amino acids to proteins.”
“The number of applications has exploded, as has the number of generative AI companies. We are moving technology at a rate four times every year over the course of 10 years. We continue to find that AI continues to become smarter. After you are done training the model, when you use ChatGPT, you write a prompt instead of programming. The AI processes through a large neural network and produces a sequence of words.”
Read more: My version of NVIDIA is ‘Vidya’, RIL’s Mukesh Ambani to NVIDIA CEO Jensen Huang
“We realized that intelligence is not just one-shot; it requires thinking. Thinking results in higher-quality answers. We have discovered a second scaling model: the longer you think, the higher-quality answers you can produce. There are some problems, like ‘What is my favorite Indian food?’ It’s chicken biryani. I will tell you, I don’t have to reason about that.”
“If I want to travel from California to Mumbai with stops in other cities, and I tell ChatGPT that I would like to do this within 3 days, mentioning hotels and my desired departure times, the number of permutations is quite high. Planning this process is very complicated. That’s where thinking, reasoning, and planning come in. We now have two fundamentally scaling laws: the first is training, and now it is inference. We are going to deliver Blackwells by Q4. The demand for Blackwells is incredible. The number of foundational models has more than doubled. More and more companies are realizing that fundamental intelligence is vital to their company.”
“The size of the models has increased by 20-30x. Blackwells is also used for generating tokens at incredible speed.”
“AI models will help our employees become super employees,” says Huang and adds, “We’re gonna have agents or agentic AI models that will help our employees become super employees and more productive. Working with our partners to integrate libraries into their platforms so that they could enable agents to be created, on-boarded, deployed, and improved into a life cycle of agents.”
AI and Arts: Dash of Bollywood magic
While talking to Akshay Kumar, Huang highlighted their Thailand connection and said, “One of us is a martial artist and one of us has 80 million followers. Akshay went to Thailand to become a martial artist, I went to Thailand to grow up.”
“While talking about what is one thing which AI cannot copy from humans, Huang said, ‘AI has no possibility of doing all of what we do. However, depending on our jobs, sometimes it can do 20 percent of our jobs 1,000 times better, but in no job, can they do all of it. Because of that, everyone of us should apply AI to automate the 20/40/50 percent. The person who uses AI to automate that 20 percent is going to take your job; it’s not AI.”
Ambani and Huang
While talking to Mukesh Ambani, Huang said, “No one has contributed more to help India become more high tech and deep tech than you. You have deep aspirations to make India a deep tech industry.”
Ambani said, “When you are in India, I have got to give you my version of what NVIDIA means to me. Vidya is a very important word. In Hindi, it means knowledge.”
“Everybody said Nvidia is a horrible name, you will never make it,” says Huang, when Mukesh Ambani tells him vidya means knowledge in Sanskrit. Huang jokingly takes credit, saying, “I knew it! I knew I named Nvidia right… 22 years ago.”
“We are at the doorstep of the new intelligence age. Welcome to India. Thank you for actually contributing to the world, bringing the intelligence age into our lifetime. Hopefully, together with everyone, this can drive more prosperity to the world and India,” said Ambani.
“Very few countries in the world has this natural resource called IT, like India,” said Huang.
“Jio is the largest data company in the world,’ said Ambani.
“As our PM has said that this is the new aspirational India. Only country in the world with average age of 1.2 bn people is below 35,” said Mukesh Ambani
“India has become home among all companies across world– we are fastest growing across industries.”
“India is fast becoming innovation hub for the world.”
“Apart from the US and China, India has the best digital connectivity infrastructure.”
“We also have the necessary infrastructure. We are fortunate to have the connectivity infrastructure.”
“India will be a intelligence market.”
“We will surprise the world with what India and Indian can do in the artificial intelligence market.”
“Indians will not only export CEOs to the worlds largest companies but also millions of Indian who will deliver AI services to the world. Everyone will have to work together to bring the intelligence age safely to the world.”
On PM Modi and Jobs
“It makes complete sense that India should manufacture its own AI. You should not export data to import intelligence. India should not export flour to import bread,” says Jensen Huang on PM Modi. He also said that the Prime Minister asked him to address his cabinet about artificial intelligence.